Factor-Augmented Autoregressive Neural Network to forecast NOx in the city of Madrid
Air pollution poses a significant threat to public health and the environment in urban areas worldwide. In the context of urban air quality, nitrogen oxides (NOx), comprising nitrogen dioxide (NO2) and nitric oxide (NO), stand out as key pollutants with well-documented adverse effects. The city of M...
| Autores: | , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2024 |
| País: | España |
| Institución: | Universidad de Castilla-La Mancha |
| Repositorio: | RUIdeRA. Repositorio Institucional de la UCLM |
| OAI Identifier: | oai:ruidera.uclm.es:10578/40531 |
| Acceso en línea: | https://www.sciencedirect.com/science/article/pii/S0038012124001575 https://hdl.handle.net/10578/40531 |
| Access Level: | acceso abierto |
| Palabra clave: | Environmental forecasting Factor model Neural networks Principal Component Spatio-temporal data |
| Sumario: | Air pollution poses a significant threat to public health and the environment in urban areas worldwide. In the context of urban air quality, nitrogen oxides (NOx), comprising nitrogen dioxide (NO2) and nitric oxide (NO), stand out as key pollutants with well-documented adverse effects. The city of Madrid, as the capital and largest urban center of Spain and the third largest of Europe, is no exception to the challenges posed by NOx pollution. Most of the recent literature on forecasting air pollution, and specifically on NOx, is based on the use of Neural Networks (NN). Little is known about the forecasting ability of factor models in this context. The main aim of this paper is to use Factor-Augmented Autoregressive Neural Networks (FA-ARNN-X) to predict future patterns of NOx pollutants in the territorial monitoring stations of Madrid, using lagged NOx values, meteorological variables and latent factors. The main results indicate that the proposed forecasting model provides statistically more accurate predictions of air pollution than its competing benchmarks and should be used by policymakers for more accurate air pollution monitoring. |
|---|